Table 3.
Overview of the Aims and Settings of the Following Simulation Studies.
Study | Central question | Experimental factors | Criterion variables |
---|---|---|---|
I | Are the methods mislead by actual differences in the person parameter distributions (PPD)? | DIF: No/Yes Mean difference in PPD: No/Yes Variance difference in PPD: No/Yes Covariate pattern: Binary/Numeric |
Type I error/Power |
II | Are the methods sensitive to DIF in groups formed by complex covariate patterns? | DIF: No/Yes Covariate pattern: Categorical/Numeric-80/U-shaped/Interaction |
Type I error/power |
III | Are the methods also sensitive to DSF? | DSF pattern: Single-level/Convergent/Divergent/Balanced | Power |
Note. DIF = differential item functioning; DSF = differential step functioning.